Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: moderated mediation logistic outcome

From   Richard Williams <>
Subject   RE: st: moderated mediation logistic outcome
Date   Fri, 28 Oct 2011 12:17:08 -0500

At 10:37 AM 10/28/2011, Cameron McIntosh wrote:
I don't know about the precise Stata code for estimating a mod-med model, but I would caution you that if you are using a series of logistic regression steps to test for the presence of moderated mediation, you will need to account for the fact that the variance of the outcome depends on the model (and will change across equations as predictors and interactions terms are added) -- all this due to the identification constraint that error = pi-squared/3 = 3.29. This also means that coefficients will be recaled to respect the constraint, and will thus need to be standardized to allow for comparisons across equations. Note that the same issue applies to the probit case with error = 1.0. You may wish to see:

You may also wish to look at the -khb- program (available from SSC) and the papers related to it. You may need to write the authors to get copies of the papers, as I believe they have 2 pieces that are currently forthcoming and some others that are in progress.

Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index